xref: /titanic_51/usr/src/man/man1/truss.1 (revision bbaa8b60dd95d714741fc474adad3cf710ef4efd)
te
Copyright (c) 2007, Sun Microsystems, Inc. All Rights Reserved.
Copyright 1989 AT&T
The contents of this file are subject to the terms of the Common Development and Distribution License (the "License"). You may not use this file except in compliance with the License.
You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE or http://www.opensolaris.org/os/licensing. See the License for the specific language governing permissions and limitations under the License.
When distributing Covered Code, include this CDDL HEADER in each file and include the License file at usr/src/OPENSOLARIS.LICENSE. If applicable, add the following below this CDDL HEADER, with the fields enclosed by brackets "[]" replaced with your own identifying information: Portions Copyright [yyyy] [name of copyright owner]
TRUSS 1 "Jul 31, 2004"
NAME
truss - trace system calls and signals
SYNOPSIS

truss [-fcaeildDE] [- [tTvx] [!] syscall ,...]
 [- [sS] [!] signal ,...] [- [mM] [!] fault ,...]
 [- [rw] [!] fd ,...]
 [- [uU] [!] lib ,... : [:] [!] func ,...]
 [-o outfile] command | -p pid[/lwps]...
DESCRIPTION

The truss utility executes the specified command and produces a trace of the system calls it performs, the signals it receives, and the machine faults it incurs. Each line of the trace output reports either the fault or signal name or the system call name with its arguments and return value(s). System call arguments are displayed symbolically when possible using defines from relevant system headers. For any path name pointer argument, the pointed-to string is displayed. Error returns are reported using the error code names described in Intro(3). If, in the case of an error, the kernel reports a missing privilege, a privilege name as described in privileges(5) is reported in square brackets ([ ]) after the error code name.

Optionally (see the -u option), truss also produce an entry/exit trace of user-level function calls executed by the traced process, indented to indicate nesting.

OPTIONS

For those options that take a list argument, the name all can be used as a shorthand to specify all possible members of the list. If the list begins with a !, the meaning of the option is negated (for example, exclude rather than trace). Multiple occurrences of the same option can be specified. For the same name in a list, subsequent options (those to the right) override previous ones (those to the left).

The following options are supported: -a

Shows the argument strings that are passed in each exec() system call.

-c

Counts traced system calls, faults, and signals rather than displaying the trace line-by-line. A summary report is produced after the traced command terminates or when truss is interrupted. If -f is also specified, the counts include all traced system calls, faults, and signals for child processes.

-d

Includes a time stamp on each line of trace output. The time stamp appears as a field containing seconds\|.\|fraction at the start of the line. This represents a time in seconds relative to the beginning of the trace. The first line of the trace output shows the base time from which the individual time stamps are measured, both as seconds since the epoch (see time(2)) and as a date string (see ctime(3C) and date(1)). The times that are reported are the times that the event in question occurred. For all system calls, the event is the completion of the system call, not the start of the system call.

-D

Includes a time delta on each line of trace output. The value appears as a field containing seconds\|.\|fraction and represents the elapsed time for the LWP that incurred the event since the last reported event incurred by that LWP. Specifically, for system calls, this is not the time spent within the system call.

-e

Shows the environment strings that are passed in each exec() system call.

-E

Includes a time delta on each line of trace output. The value appears as a field containing seconds.fraction and represents the difference in time elapsed between the beginning and end of a system call. In contrast to the -D option, this is the amount of time spent within the system call.

-f

Follows all children created by fork() or vfork() and includes their signals, faults, and system calls in the trace output. Normally, only the first-level command or process is traced. When -f is specified, the process-id is included with each line of trace output to indicate which process executed the system call or received the signal.

-i

Does not display interruptible sleeping system calls. Certain system calls, such as open() and read() on terminal devices or pipes, can sleep for indefinite periods and are interruptible. Normally, truss reports such sleeping system calls if they remain asleep for more than one second. The system call is reported again a second time when it completes. The -i option causes such system calls to be reported only once, when they complete.

-l

Includes the id of the responsible lightweight process (LWP) with each line of trace output. If -f is also specified, both the process-id and the LWP-id are included.

-m [!]fault,...

Machine faults to trace or exclude. Those faults specified in the comma-separated list are traced. Faults can be specified by name or number (see <sys/fault.h>). If the list begins with a !, the specified faults are excluded from the trace output. Default is -mall -m !fltpage.

-M [!]fault,...

Machine faults that stop the process. The specified faults are added to the set specified by -m. If one of the specified faults is incurred, truss leaves the process stopped and abandoned (see the -T option). Default is -M!all.

-o outfile

File to be used for the trace output. By default, the output goes to standard error.

-p

Interprets the command arguments to truss as a list of process-ids for existing processes (see ps(1)) rather than as a command to be executed. truss takes control of each process and begins tracing it provided that the userid and groupid of the process match those of the user or that the user is a privileged user. Users can trace only selected threads by appending /thread-id to the process-id. Mutiple threads can be selected using the - and , delimiters. For example /1,2,7-9 traces threads 1, 2, 7, 8, and 9. Processes can also be specified by their names in the /proc directory, for example, /proc/12345.

-r [!]fd,...

Shows the full contents of the I/O buffer for each read() on any of the specified file descriptors. The output is formatted 32 bytes per line and shows each byte as an ASCII character (preceded by one blank) or as a 2-character C language escape sequence for control characters such as horizontal tab (\|\e\|t) and newline (\|\e\|n). If ASCII interpretation is not possible, the byte is shown in 2-character hexadecimal representation. (The first 12 bytes of the I/O buffer for each traced print >read() are shown even in the absence of -r.) Default is -r!all.

-s [!]signal,...

Signals to trace or exclude. Those signals specified in the comma-separated list are traced. The trace output reports the receipt of each specified signal, even if the signal is being ignored (not blocked). (Blocked signals are not received until they are unblocked.) Signals can be specified by name or number (see <sys/signal.h>). If the list begins with a !, the specified signals are excluded from the trace output. Default is -sall.

-S [!]signal,...

Signals that stop the process. The specified signals are added to the set specified by -s. If one of the specified signals is received, truss leaves the process stopped and abandoned (see the -T option). Default is -S!all.

-t [!]syscall,...

System calls to trace or exclude. Those system calls specified in the comma-separated list are traced. If the list begins with a !, the specified system calls are excluded from the trace output. Default is -tall.

-T [!]syscall,...

Specifies system calls that stop the process. The specified system calls are added to the set specified by -t. If one of the specified system calls is encountered, truss leaves the process stopped and abandoned. That is, truss releases the process and exits but leaves the process in the stopped state at completion of the system call in question. A debugger or other process inspection tool (see proc(1)) can then be applied to the stopped process. truss can be reapplied to the stopped process with the same or different options to continue tracing. Default is -T!all. A process left stopped in this manner cannot be restarted by the application of kill -CONT because it is stopped on an event of interest via /proc, not by the default action of a stopping signal (see signal.h(3HEAD)). The prun(1) command described in proc(1) can be used to set the stopped process running again.

-u [!]lib,...:[:][!]func,\|.\|.\|.

User-level function call tracing. lib,\|.\|.\|. is a comma-separated list of dynamic library names, excluding the ``.so.n'' suffix. func,\|.\|.\|. is a comma-separated list of function names. In both cases the names can include name-matching metacharacters *,?,[] with the same meanings as those of sh(1) but as applied to the library/function name spaces, not to files. An empty library or function list defaults to *, trace all libraries or functions in a library. A leading ! on either list specifies an exclusion list, names of libraries or functions not to be traced. Excluding a library excludes all functions in that library; any function list following a library exclusion list is ignored. A single : separating the library list from the function list means to trace calls into the libraries from outside the libraries, but omit calls made to functions in a library from other functions in the same library. A double :\|: means to trace all calls, regardless of origin. Library patterns do not match either the executable file or the dynamic linker unless there is an exact match (l* does not match ld.so.1). To trace functions in either of these objects, the names must be specified exactly, as in:

truss -u a.out -u ld ...
a.out is the literal name to be used for this purpose; it does not stand for the name of the executable file. Tracing a.out function calls implies all calls (default is ::). Multiple -u options can be specified and they are honored left-to-right. The id of the thread that performed the function call is included in the trace output for the call. truss searches the dynamic symbol table in each library to find function names and also searches the standard symbol table if it has not been stripped.
-U [!]lib,\|.\|.\|.\|:[:][!]func,\|.\|.\|.

User-level function calls that stop the process. The specified functions are added to the set specified by -u. If one of the specified functions is called, truss leaves the process stopped and abandoned (see the -T option).

-v [!]syscall,...

Verbose. Displays the contents of any structures passed by address to the specified system calls (if traced by -t). Input values as well as values returned by the operating system are shown. For any field used as both input and output, only the output value is shown. Default is -v!all.

-w [!]fd,...

Shows the contents of the I/O buffer for each write() on any of the specified file descriptors (see the -r option). Default is -w!all.

-x [!]syscall,...

Displays the arguments to the specified system calls (if traced by -t) in raw form, usually hexadecimal, rather than symbolically. This is for unredeemed hackers who must see the raw bits to be happy. Default is -x!all.

See man pages section 2: System Calls for system call names accepted by the -t, -T, -v, and -x options. System call numbers are also accepted.

If truss is used to initiate and trace a specified command and if the -o option is used or if standard error is redirected to a non-terminal file, then truss runs with hangup, interrupt, and quit signals ignored. This facilitates tracing of interactive programs that catch interrupt and quit signals from the terminal.

If the trace output remains directed to the terminal, or if existing processes are traced (the -p option), then truss responds to hangup, interrupt, and quit signals by releasing all traced processes and exiting. This enables the user to terminate excessive trace output and to release previously-existing processes. Released processes continue normally, as though they had never been touched.

EXAMPLES

Example 1 Tracing a Command

The following example produces a trace of the find(1) command on the terminal:

example$ truss find . -print >find.out

Example 2 Tracing Common System Calls

The following example shows only a trace of the open, close, read, and write system calls:

example$ truss -t open,close,read,write find . -print >find.out

Example 3 Tracing a Shell Script

The following example produces a trace of the spell(1) command on the file truss.out:

example$ truss -f -o truss.out spell document

spell is a shell script, so the -f flag is needed to trace not only the shell but also the processes created by the shell. (The spell script runs a pipeline of eight processes.)

Example 4 Abbreviating Output

The following example abreviates output:

example$ truss nroff -mm document >nroff.out

because 97% of the output reports lseek(), read(), and write() system calls. To abbreviate it:

example$ truss -t !lseek,read,write nroff -mm document >nroff.out

Example 5 Tracing Library Calls From Outside the C Library

The following example traces all user-level calls made to any function in the C library from outside the C library:

example$ truss -u libc ...

Example 6 Tracing library calls from within the C library

The following example includes calls made to functions in the C library from within the C library itself:

example$ truss -u libc:: ...

Example 7 Tracing Library Calls Other Than the C Library

The following example traces all user-level calls made to any library other than the C library:

example$ truss -u '*' -u !libc ...

Example 8 Tracing printf and scanf Function Calls

The following example traces all user-level calls to functions in the printf and scanf family contained in the C library:

example$ truss -u 'libc:*printf,*scanf' ...

Example 9 Tracing Every User-level Function Call

The following example traces every user-level function call from anywhere to anywhere:

example$ truss -u a.out -u ld:: -u :: ...

Example 10 Tracing a System Call Verbosely

The following example verbosely traces the system call activity of process #1, init(1M) (if you are a privileged user):

example# truss -p -v all 1

Interrupting truss returns init to normal operation.

FILES
/proc/*

Process files

SEE ALSO

date(1), find(1), proc(1), ps(1), sh(1), spell(1), init(1M), Intro(3), exec(2), fork(2), lseek(2), open(2), read(2), time(2), vfork(2), write(2), ctime(3C), signal.h(3HEAD), proc(4), attributes(5), privileges(5), threads(5)

man pages section 2: System Calls

NOTES

Some of the system calls described in man pages section 2: System Calls differ from the actual operating system interfaces. Do not be surprised by minor deviations of the trace output from the descriptions in that document.

Every machine fault (except a page fault) results in the posting of a signal to the LWP that incurred the fault. A report of a received signal immediately follows each report of a machine fault (except a page fault) unless that signal is being blocked.

The operating system enforces certain security restrictions on the tracing of processes. In particular, any command whose object file (a.out) cannot be read by a user cannot be traced by that user; set-uid and set-gid commands can be traced only by a privileged user. Unless it is run by a privileged user, truss loses control of any process that performs an exec() of a set-id or unreadable object file; such processes continue normally, though independently of truss, from the point of the exec().

To avoid collisions with other controlling processes, truss does not trace a process that it detects is being controlled by another process via the /proc interface. This allows truss to be applied to proc(4)-based debuggers as well as to another instance of itself.

The trace output contains tab characters under the assumption that standard tab stops are set (every eight positions).

The trace output for multiple processes or for a multithreaded process (one that contains more than one LWP) is not produced in strict time order. For example, a read() on a pipe can be reported before the corresponding write(). For any one LWP (a traditional process contains only one), the output is strictly time-ordered.

When tracing more than one process, truss runs as one controlling process for each process being traced. For the example of the spell command shown above, spell itself uses 9 process slots, one for the shell and 8 for the 8-member pipeline, while truss adds another 9 processes, for a total of 18.

Not all possible structures passed in all possible system calls are displayed under the -v option.